(This is a sample cover image for this issue. The actual cover is not yet available at this time.)

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Profiling of oxidized lipid products of marine fish under acute oxidative stress

Ming Long Sirius Chung, Kai Yan Eric Lee, Chung-Yung Jetty Lee ⇑School of Biological Sciences, The University of Hong Kong, Pokfulam Road, Hong Kong

a r t i c l e i n f o

Article history:Received 12 September 2012Accepted 27 November 2012Available online 5 December 2012

Keywords:DHAEPAIsoprostanesHETECholesterol oxidation productsMedakaHydrogen peroxide

a b s t r a c t

Free radical products including reactive oxygen species are potent to oxidize lipids and reliable measure-ments have been established mostly in human and rodent. To date, robust biomarkers were not used toassess the peroxidation in marine fish. The changes of oxidized lipid products from polyunsaturated fattyacids and cholesterol were assessed after exposure of H2O2 to fish (medaka). Oxidized lipid productsreleased by free radical reaction (F2-isoprostanes and metabolites, F3-isoprostanes, neuroprostanes,7-ketocholesterol, 7b-hydroxycholesterol), by lipoxygenase enzymes (5(S)-, 8(S)-, 12(S)- and 15(S)-HETE,and resolvin D1) and by cytochrome P450 (9(S)-, 11(S)- and 20-HETE, and 27-hydroxycholestrol) weremeasured in fish muscle using LC/MS/MS. Arachidonate, docosahexaenoate, eicosapentaenoate and cho-lesterol levels, and antioxidant enzymes activity (catalase, SOD and gluthathione reductase) measure-ment were also determined. Activity of antioxidant enzymes especially catalase were elevated inpresence of H2O2 however longer exposure time suppressed the antioxidant activities. Arachidonate,docosahexaenoate, eicosapentaenoate and cholesterol levels were reduced in presence of H2O2 andoxidized lipid products (isoprostanes, neuroprostanes 5(S)-HETE, 20-HETE, 7-ketocholesterol, 27-hydrox-ycholesterol and resolvin D1) were rapidly released in the fish muscle. This study validates oxidized lipidproducts, noticeably isoprostanes are measurable in marine fish muscle and should be considered whenassessing oxidative stress especially due to exogenous factors.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Fish is an important source of polyunsaturated fatty acids(PUFA) in human diet. PUFA from fish consumption in particulardocosahexaenoic and eicosapentaenoic acids aids in preventingcancer, coronary heart disease, inflammation and neuronal devel-opment in human (Streppel et al., 2008; Quinn et al., 2010; Vander Meij et al., 2010; Skulas-Ray et al., 2011). However, these PUFAare prone to peroxidation by free radical products including reac-tive oxygen species (O��2 , �OH, H2O2) generated from natural causessuch as hypoxia (eutrophication) in seawater and man-madecauses from heavy metals and numerous organic contaminants, re-dox cycling compounds, polycyclic aromatic hydrocarbons, anddioxins (Lesser, 2006; Valavanidis et al., 2006; Halliwell andGutteridge, 2007). Reactive oxygen species are detoxified by anti-oxidant defense system, which comprises of antioxidant enzymes(superoxide dismutase, catalase and glutathione) and antioxidantscavengers. The imbalance between production of reactive oxygenspecies and antioxidant defenses lead to oxidative stress and sub-sequent damage to lipids, proteins, and DNA in vivo. Inherently,these changes in antioxidant defenses and oxidative damageare used as biomarkers of oxidative stress in biological studies(Halliwell and Gutteridge, 2007).

A wide range of oxidized lipid products can be generated inhuman, animal and cell samples (Roberts and Morrow, 2000;Kadiiska et al., 2005). Prostaglandin-like F2-isoprostanes (F2-IsoPs) are formed in vivo through non-enzymatic free radical-in-duced peroxidation of arachidonic acid and considered by manyto be the best indicators of lipid peroxidation in biological spe-cies (Roberts and Morrow, 2000; Kadiiska et al., 2005). Thereare 64 isomers of F2-IsoPs and it is shown (Roberts and Morrow,2000) the most common type, 8-iso-prostaglandin F2a canbe further metabolized via b-oxidation and subsequent reduc-tion mechanism to give 2,3-dinor-8-iso-prostaglandin F2a(F2-IsoPM1) and 2,3-dinor-5,6-dihydro-8-prostaglandin F2a(F2-IsoPM2) respectively in urine (Roberts et al., 1996). Eicosa-pentaenoic acid and docosahexaenoic acid are metabolized bycyclooxygenases to prostaglandins and transformed by non-enzymatic free radicals to give respectively 8-iso-prostaglandinF3a (F3-IsoPs) and neuroprostanes (F4-NP). Arachidonic acidin vivo can also be oxidized by lipoxygenase or cytochromeP450 enzymes to release hydroxyeicosatetraenoic acid products(HETEs). Likewise docosahexaenoic acid can be oxidized bylipoxygenase to release protectin D1 and resolvin D1. Oxidationproducts can also arise from cholesterol (COPs). It can be oxi-dized by enzymatic (cytochrome P450-dependent) reactions torelease 24- and 27-hydroxycholesterol, and 7a-hydroxycholes-terol or by oxidative damage to release 7b-hydroxycholesteroland 7-ketocholesterol (Garenc et al., 2010).

0278-6915/$ - see front matter � 2012 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.fct.2012.11.047

⇑ Corresponding author. Fax: +852 2559 9114.E-mail address: jettylee@hku.hk (C.-Y.J. Lee).

Food and Chemical Toxicology 53 (2013) 205–213

Contents lists available at SciVerse ScienceDirect

Food and Chemical Toxicology

journal homepage: www.elsevier .com/locate / foodchemtox

Author's personal copy

Many studies have used measurements of antioxidant enzymesor malondialdehyde levels or thiobarbituric reactive species(TBARS) to determine oxidative stress level in fish (Tsangariset al., 2011a, 2011b). These oxidized lipid product biomarkers arenon-specific and non-reliable indicator of oxidative stress (Knightet al., 1988). So far studies are scarce in examining oxidative dam-age of marine fish using multiple biomarkers by more reliablemethod such as gas chromatography/mass spectrometry (GC/MS)or liquid chromatography tandem mass spectrometry (LC/MS/MS) (Basu, 1999; Roberts and Morrow, 2000; Lee et al., 2004). Moststudies that assess the reliability of oxidized lipid products havebeen performed in animals in particular rodents (Basu, 1999;Kadiiska et al., 2005) and humans (Lee et al., 2004, 2009; Limet al., 2010). Biomarkers of oxidized lipid products in marine fishmight perhaps differ from those measured in rodents and human.

The aims of this study are to assess the changes of oxidized lipidproducts after acute exposure to an oxidant (hydrogen peroxide) inmarine fish (medaka) and relate these products with the fatty acidand cholesterol responsible. Medaka (Oryzias latipes) is widely usedas laboratory model to study developmental biology and geneticsrelated to toxicology, endocrine disruptors and carcinogenesisresearch.

2. Materials and methods

2.1. Treatment of medaka fish

Adult marine medaka fish (O. latipes) was maintained in fully aerated water setat photoperiod of 14 h light/10 h dark, water temperature at 26 ± 1 �C, ammonia le-vel (0.010 ± 0.005 mg/L) and pH 8. The fishes were fed twice daily with frozen art-emia and hormone-free flake food (AX5, Aquatic Eco-Systems Inc., Florida, USA)

Fig. 1. Chromatograms of HETEs, resolvin D1, isoprostanes and neuroprostanes in lipid extract of male fish muscle after hydrogen peroxide treatment.

206 M.L.S. Chung et al. / Food and Chemical Toxicology 53 (2013) 205–213

Author's personal copy

prior to exposure to oxidants, and the water was cleaned (50% changed) every 72 h.Medaka fish (n = 72, 1:1 male to female) was treated by placing individual fish in abeaker of 500 ml marine water only, marine water made up with 200 lM or marinewater made up with 1000 lM hydrogen peroxide (H2O2) for 2 or 6 h.

2.2. Sample preparation for lipid extraction

After treatment with or without hydrogen peroxide, each medaka fish wasdecapitated and internal organs were removed. The body muscle (0.8 g) washomogenized in phosphate buffered saline (PBS, pH 7.4) and Folch solution withbutylated hydroxytoluene. After centrifugation at 800g for 10 min, at 4 �C the lowerorganic layer was transferred to a glass vial and the solvent finally evaporated un-der a stream of nitrogen gas. The dried Folch extract was hydrolyzed with PBS andpotassium hydroxide in methanol (1:1) together with heavy isotopic internal stan-dards [18O2] F4-NP (gift by Dr. Ginger Milne (Eicosanoid Core Laboratory, VanderbiltUniversity, Nashville, TN, USA), and 8-iso-PGF2a-d4, PGF2a-d4, 5(S)-HETE-d8, 12(S)-HETE-d8, 15(S)-HETE-d8, 20-HETE-d6, arachidonic acid-d8, eicosapentaenoic acid-d5

and docosahexaenoic acid-d5 (Cayman Chemicals, MI, USA). Heavy isotopic COPsinternal standards 7a-hydroxycholesterol-d7, 7b-hydroxycholesterol-d7, 27-hydroxycholesterol-d6, 7-ketocholesterol-d7 and cholesterol-d7 (Avanti Polar Lipids,USA) were also added. After hydrolysis the sample was cooled and then neutralizedby hydrochloric acid. Finally formic acid (pH 4.5) was added, mixed and thencentrifuged to remove any protein precipitate before solid phase extraction (Leeet al., 2008). For solid phase extraction 60 mg MAX columns (Oasis, Waters, USA)was used to purify the hydrolyzed samples according to Lee et al. (2009) method.

2.3. Analysis of oxidized lipid products of fatty acids

Oxidized lipid products and fatty acids were performed using 1290 Infinity LCsystem (Agilent, USA) according to (Mallat et al., 1999; Liang et al., 2003;Unterwurzacher et al., 2008; Fahy et al., 2009) with modification. Hillic C18 column(2.4 mm particle size, 50 � 2.4 mm, Phenomenex, USA) fitted with a guard columnpacked with 2.4 mm ODS material (Phenomenex, USA) was used and maintained at30 �C. Mobile phase consisted of 0.1% acetic acid in water (A) and 0.1% acetic acid inmethanol (B) and was set to 100 ll/min flow rate. The mobile phase ran in a lineargradient from phase A 50% to phase B 50% over 3 min, B 95% to A 5% for 0.5 min andB 55% to A 45% for 3 min. The sample injection volume was 10 ll.

A QTrap 3200 triple quadrupole mass spectrometer (Sciex Applied Biosystems,USA) was operated in the negative atmospheric pressure chemical ionization (APCI)mode with a source temperature of 500 �C and capillary temperature of 250 �C. Thedeclustering and entrance potential were set at �50 v and �10 V, respectively. Thecollision energy was set to �30 v and was optimized to maximize the ion currentsof the precursor to product ion dissociation. The spray voltage was set to �4000 vand nitrogen gas was used as curtain gas. The analytes were detected by MS/MSusing multiple reactions monitoring (MRM) and monitoring according to variousreports (Mallat et al., 1999; Liang et al., 2003; Unterwurzacher et al., 2008; Fahyet al., 2009).

Quantitation of each analyte was achieved by relating the peak area with itscorresponding internal standard peak. Where available, heavy labeled isotopewas used for quantitation whereas those without heavy labeled isotope, quantita-tion was achieved by using 5(S)-HETE-d8 for 8(S)-, 9(S)- and 11(S) HETE. Theresponse was 2: 1 for 5(S)-HETE-d8 to 8(S)-HETE and 9(S)-HETE, and 1: 4 for5(S)-HETE-d8 to 11(S)-HETE at equal concentration. Likewise, for 2,3-dinor-8-iso-prostaglandin F2a (F2-IsoPM1), 2,3-dinor-5,6-dihydro-8-iso-prostaglandin F2a(F2-IsoPM2) and 8-iso-prostaglandin F3a (F3-IsoPs) heavy labeled isotope 8-iso-prostaglandin F2a-d4 was used for quantification and the response was 1:1 (F2-IsoPMand F2-IsoPM2) and 1:0.8 (F3-IsoPs) respectively at equal concentration. Docosahex-aenoic acid-d5 was used for resolvin D1 and the response was 1:2 respectively atequal concentration.

2.4. Analysis of cholesterol oxidation products

Analysis by LC/MS/MS was achieved according to Shui et al. (2011) report withmodification. HPLC analysis was operated by mobile phase consisted of 5 mMammonium acetate in 100% methanol (A) and 5 mM ammonium acetate in 85%methanol (B) and was set to 200 ll/min flow rate. The mobile phase ran from 0%A phase to 100% B phase over 4 min and then B to A from 100% to 0% for 4 min.The sample injection volume was 10 ll.

The mass spectrometer (QTrap 3200, Sciex Applied Biosystems, USA) was oper-ated in the positive atmospheric pressure chemical ionization (APCI) mode with asource temperature of 400 �C and capillary temperature of 250 �C. The declusteringand entrance potential were set at 15 and 10 V, respectively. The collision energywas set to 55 v and was optimized to maximize the ion currents of the precursorto product ion dissociation. The spray voltage was set to 5500 v and nitrogen gaswas used as curtain gas. The analytes were detected by MS/MS using MRM andmonitoring the transitions as suggested by Shui et al. (2011). Quantitation of eachanalyte was achieved by relating the peak area with its corresponding internal stan-dard peak.

2.5. Analysis of antioxidant enzymes

All samples were measured for antioxidant enzyme activity (catalase, glutathi-one and superoxide dismutase) using Cayman Chemical (USA) kits. Prior to analysis,the tissue samples were extracted in cold buffer suggested. All extracts were takenand stored at �80 �C for antioxidant activities measurement using spectrophotom-eter (Hitachi, Japan) at 540 nm for catalase, 405/414 nm for glutathione reductaseand 440/460 nm for superoxide dismutase.

2.6. Statistical analysis

Statistical analysis was performed using by GraphPad Prism version 5.0 forMacintosh (GraphPad Software, CA USA). All values are expressed as mean ± SD.Student’s t-test was performed between post-dose time points and baseline. Anysignificant changes found by Student’s t-test were confirmed by analysis of variance(mixed model) for the effect of hydrogen peroxide over time. Only significance atlevel of p < 0.05 was considered in the analysis.

3. Results

All samples from the fish tissue showed the presence of isopros-tanes, neuroprostanes, HETEs COPs and fatty acids released fromthe extracted lipids by saponification. The sensitivity of the mea-surement complied with (Unterwurzacher et al., 2008) report.The chromatography elution time of the analyte ranged 0.5–1.0 min (Fig. 1).

3.1. Baseline level

The baseline level of medaka fish is shown in Table 1. A largedisparity in the baseline level is displayed between male and fe-male medaka, in particular 2,3-dinor-5,6-dihydro-8-iso-PGF2a(F2-IsoPM2) which is approximately 3 times higher in malemedaka than female medaka. Levels of 2,3-dinor-8-iso-PGF2a

Table 1Baseline level of fatty acids, prostaglandin F2a, oxidized lipid products andantioxidant enzymes of male and female medaka.

Male Female

Arachidonate 0.83 ± 0.04 3.15 ± 0.16⁄⁄

Eicosapentaenoate 49.22 ± 5.82 54.91±.9.0Docosahexaenoate 82.10 ± 6.52 48.19 ± 5.59⁄⁄

Cholesterola 4.19 ± 0.50 7.05 ± 0.77⁄⁄

Non-enyzmatic oxidationF2-IsoPs 1.88 ± 0.11 3.18 ± 0.21⁄⁄

2,3-Dinor-8-iso-PGF2a 2.40 ± 0.37 9.70 ± 0.19⁄⁄

2,3-Dinor-5,6-dihydro-8-iso-PGF2a 119.80 ± 9.47 50.0 ± 6.22⁄⁄

F3-IsoPs 0.86 ± 0.07 2.46 ± 0.26⁄⁄

F4-NP 0.84 ± 0.18 0.55 ± 0.06⁄⁄

7b-Hydroxycholesterola 2.43 ± 0.19 4.06 ± 0.31⁄⁄

7-Ketocholesterola 1.09 ± 0.16 1.32 ± 0.15⁄

Enzymatic oxidationLipoxygenase mediated5(S)-HETE 18.07 ± 2.27 8.04 ± 1.24⁄⁄

8(S)-HETE 4.43 ± 1.10 1.44 ± 0.18⁄⁄

12(S)-HETE 29.81 ± 4.04 4.18 ± 0.62⁄⁄

15(S)-HETE 1.15 ± 0.21 1.34 ± 0.19⁄

Resolvin D1 0.29 ± 0.09 1.91 ± 0.21⁄⁄

Cytochrome mediated9(S)-HETE 3.99 ± 0.86 1.45 ± 0.23⁄⁄

11(S)-HETE 7.46 ± 0.62 2.60 ± 0.12⁄⁄

20-HETE 0.39 ± 0.07 0.42 ± 0.0427-Hydroxycholesterola 0.32 ± 0.05 0.30 ± 0.03

Antioxidant enzymesCatalaseb 0.82 ± 0.57 4.82 ± 1.19⁄⁄

Superoxide dismutaseb 23.44 ± 14.60 9.08 ± 3.77⁄⁄

Gluthathione reductaseb 59.98 ± 12.80 85.55 ± 28.56⁄⁄

All values expressed as mean ± SD ng/g tissue (n = 3) except those annotated a

mean ± SD lg/g tissue and b mean ± SD Unit/g tissue for catalase and superoxidedismutase, and lM/g tissue for glutathione reductase. Significant levels betweenmale and female medaka are indicated by ⁄and ⁄⁄p < 0.05 and p < 0.01, respectively.

M.L.S. Chung et al. / Food and Chemical Toxicology 53 (2013) 205–213 207

Author's personal copy

(F2-IsoPM1), F3-IsoPs, resolvin D1, 7b-hydroxycholesterol, 7-keto-cholesterol and 15(S)-HETE were significantly lower in male meda-ka than female medaka. However F4-NP and 5(S)-, 8(S)-, 9(S)-,11(S)- and 12(S)-HETE levels were significantly higher in male me-daka than female medaka, and no difference was found betweenmale and female medaka for 20-HETE and 27-hydroxycholesterollevels (Table 1). Lipid substrates, arachidonate and cholesterol lev-els were lower in male medaka than female medaka, docosahexa-enoate level was higher in male medaka than female medaka andno significant difference was found between male and female me-daka for eicosapentaenoate levels.

Antioxidant activities were also found to differ between maleand female medaka. Female medaka had significantly high(p < 0.01) catalase and glutathione reductase activities but lowsuperoxide dismutase activity (p < 0.01) compared to malemedaka.

3.2. Fatty acid and cholesterol

Arachidonate levels significantly reduced (p < 0.01) in male andfemale medaka with increasing exposure time at both H2O2 con-centrations (Fig. 2A). Increasing H2O2 concentration from 200 to1000 lM augmented arachidonate level (p < 0.01) in male medakabut reduced in female medaka (p < 0.01). Eicosapentaenoate levelsignificantly reduced (p < 0.01) in male medaka with increasingexposure time at 1000 lM H2O2 (Fig. 2B), while it reduced(p < 0.01) at both H2O2 concentrations in female medaka. StrongerH2O2 concentration (1000 lM) decreased eicosapentaenoate level

(p < 0.01) in male and female medaka (p < 0.01). Increasing expo-sure time of H2O2 to male and female medaka decreaseddocosahexaenoate levels (p < 0.01). In addition, increasing H2O2

concentration from 200 to 1000 lM further reduced docosahexae-noate level (p < 0.01) in female medaka (Fig. 2C). Cholesterol levelalso reduced (p < 0.01) with increased H2O2 exposure time in maleand female medaka and suppressed in male medaka when the con-centration increased from 200 to 1000 lM (Fig. 2D).

3.3. Peroxidation of lipids: direct effect of reactive oxygen species

Reactive oxygen species and free radicals are capable of induc-ing lipid peroxidation independently. F2-IsoPs and metabolites (F2-IsoPM1 and F2-IsoPM2), F3-IsoPs, F4-NP, 7b-hydroxycholesteroland 7-ketocholesterol are some of the products generated by thereactive oxygen species and free radicals acting on polyunsatu-rated fatty acids (especially arachidonic, eicosapentaenoic anddocosahexaenoic acids) and cholesterol. As shown in Fig. 3, thepresence of H2O2 in general significantly induced the oxidized lipidproducts (Fig. 3A–E, G) except 7b-hydroxycholesterol (Fig. 3F) thatwas significantly reduced.

Increasing H2O2 from 200 to 1000 lM did not necessarilyincrease all oxidized lipid products. Our observation showedF2-IsoPs, F3-IsoPs and F4-NP levels augmented with increasedH2O2 concentration in female medaka and not so in malemedaka (Fig. 3A, D and E). When the exposure time was increasedfrom 2 to 6 h, only in female medaka significantly increase inF2-IsoPM2 level with increasing exposure time and increased

Fig. 2. Levels of arachidonate, eicosapenataenoate, docosahexaenoate and cholesterol of male and female medaka with or without hydrogen peroxide exposure. Significantdifference ⁄p < 0.05 and ⁄⁄p < 0.01 by analysis of variance between time points. All bar column values are mean ± SD, n = 3. Significant difference between 200 and 1000 lM atrespective time are annotated by +p < 0.05 and ++p < 0.01.

208 M.L.S. Chung et al. / Food and Chemical Toxicology 53 (2013) 205–213

Author's personal copy

H2O2 concentration from 200 to 1000 lM. Also F4-NP levelselevated (p < 0.01) in female medaka at 1000 lM H2O2 concentra-tion when exposure time increased from 2 to 6 h (Fig. 3E).

3.4. Peroxidation of lipids: regulated by lipoxygenase enzyme

Lipids may be oxidized by the action of lipoxygenase enzyme toproduce 5(S)-, 8(S)-, 12(S)- and 15(S)-HETE from arachidonic acid,and resolvin D1 from docosahexaenoic acid. All of these products

were significantly elevated when exposed to H2O2. Levels of 5(S)-and 8(S)-HETE also elevated with increased H2O2 exposure timeat 200 and 1000 lM H2O2 in male and female medaka. In addition,5(S)-HETE level further elevated (p < 0.01) in male medaka whenH2O2 concentration increased from 200 to 1000 lM (Fig. 4).Increasing exposure time of 1000 lM H2O2 augmented 12(S)-HETEand 15(S)-HETE levels (p < 0.01) in female medaka. These productsalso raised (p < 0.01) in female medaka when the H2O2 concentra-tion increased from 200 to 1000 lM (Fig. 4).

Fig. 3. Non-enzymatic free radical produced oxidized lipid products released by male and female medaka with or without hydrogen peroxide exposure. Significant difference⁄p < 0.05 and ⁄⁄p < 0.01 by analysis of variance between time points. All bar column values are mean ± SD, n = 3. Significant difference between 200 and 1000 lM at respectivetime are annotated by +p < 0.05 and ++p < 0.01.

M.L.S. Chung et al. / Food and Chemical Toxicology 53 (2013) 205–213 209

Author's personal copy

3.5. Peroxidation of lipids: regulated by cytochrome P450 enzyme

Oxidation of arachidonic acid can also occur via cytochromeP450 enzyme to release 9(S)- and 11(S)-HETE and 20-HETE. Levelsof these products were elevated in presence of H2O2 (Fig. 5). How-ever, 20-HETE levels increased (p < 0.01) with increased exposuretime of 1000 lM H2O2 in male and female medaka. Likewise cho-lesterol is oxidized by cytchrome P450 to release 27-hydroxycho-lesterol. In female medaka, 27-hydroxycholesterol levelsincreased (p < 0.01) with increased exposure time at 1000 lMH2O2 (Fig. 5).

3.6. Antioxidant enzyme activity

In response to oxidative stress, antioxidant enzymes will acti-vate to scavenge the reactive oxygen species. Specifically, catalasewill neutralize hydrogen peroxide and this is shown in Fig. 6A.

Presence of hydrogen peroxide elevated catalase activities in maleand female medaka. Increasing H2O2 concentration from 200 to1000 lM augmented catalase activity in male and female medakabut extended exposure time reduced the activity. Superoxide dis-mutase and glutathione reductase activities (Fig. 6B and C) weresignificantly reduced in male medaka after 6 h of 1000 lM H2O2

exposure. Superoxide dismutase activity was elevated in femalemedaka after 2 h exposure of 200 and 1000 lM H2O2 but was re-duced after 6 h (Fig. 6B). Gluthathione reductase activity was alsoreduced in female medaka after H2O2 exposure (Fig. 6C).

4. Discussion

This investigation is the first to examine multiple oxidizedlipid products in fish when challenged by oxidant. Our findingsshowed exposure of hydrogen peroxide elevated oxidized lipidproducts namely F2-IsoPs and metabolites, F3-IsoPs, F4-NP and

Fig. 4. Oxidized lipid products produced via lipoxygenase enzyme released by male and female medaka with or without hydrogen peroxide exposure. Significant difference⁄p < 0.05 and ⁄⁄p < 0.01 by analysis of variance between time points. All bar column values are mean ± SD, n = 3. Significant difference between 200 and 1000 lM at respectivetime are annotated by +p < 0.05 and ++p < 0.01.

210 M.L.S. Chung et al. / Food and Chemical Toxicology 53 (2013) 205–213

Author's personal copy

7-ketocholesterol as well as those associated to cytochrome P450enzymatic pathway (9(S)-HETE, 20-HETE and 27-hydroxycholes-terol) and those associated to lipoxygenase enzymatic pathway(5(S)-, 8-(S)-, 12(S)- and 15(S)-HETE, and resolvin D1). Manystudies (Tsangaris et al., 2011a,b) revealed oxidative stress statusin fish based on antioxidant enzymes activity and it was foundelevated in presence of oxidants in this investigation but prolongedexposure suppressed the enzymes.

In this investigation, we used 200 lM (0.68 mg/L) and 1000 lM(3.4 mg/L) hydrogen peroxide concentration to identify the changesof the oxidized lipid products. In natural seawater environment, cat-alase active microbes are abundant to maintain low concentration ofhydrogen peroxide; however, if the concentration of hydrogen per-oxide increases substantially, then the degradation rates will beslower because of the toxicity of hydrogen peroxide. The rate of deg-radation product depends on microbial population size. It is pro-posed the rate of hydrogen peroxide degradation to water andoxygen in presence of aerobic bacteria is slow at concentration below500 mg/L (8–20 days) and fast above 500 mg/L (2.5 days). Neverthe-less, it is anticipated that the physiological response of hydrogen per-oxide varies between fish species. Hydrogen peroxide baths are oftenused to treat sea lice (Lepeophtheirus and Caligus) infestations ofAtlantic salmon in seawater; the effect on growth and mortality ispoorly understood but it appears not to be affected up to 1500 mg/L hydrogen peroxide concentration. However, it appears to have det-rimental effect on the growth and development in juvenile rainbowtrout when exposed to 200 mg/L hydrogen peroxide. Regardless,hydrogen peroxide was shown to be valuable in controlling fungalinfection of fish eggs where no mortality was shown in rainbow,brown and lake trout at treatments up to 0.5–1.0 mg/L.

It can be argued that the levels of hydrogen peroxide used inthis study is excessive compared to some investigations to hydro-

gen peroxide concentration on natural surface seawater (0.001–0.0136 mg/L). The efficacy and toxicity of hydrogen peroxide inseawater varies with maximum dissolved oxygen levels and tem-perature. However base on ecotoxicology information of hydrogenperoxide 96 h LC50 for fish is approximately 1000 lM, thereforemaximum level of 1000 lM and lower level 200 lM were adaptedin this study in order to observe the extreme effect of hydrogenperoxide that leads to oxidative damage in fish in order to identifythe appropriate biomarker of oxidative stress in marine fish. Moreso, the effect of hydrogen peroxide and oxidative stress in fishbathing can be identified to understand if there is any dynamic ef-fect in lipid metabolism that indeed was observed in this study. Wefocused on one area of oxidative damage namely lipids in thisstudy as it is abundantly found in fish and an important humandietary source of omega-3 and omega-6 fatty acid. Consumptionof these fatty acids can help in neuronal development, and reduceinflammation, risk of coronary heart disease and cancer (Streppelet al., 2008; Quinn et al., 2010; Van der Meij et al., 2010; Skulas-Rayet al., 2011). We also used a reliable means of measuring oxidativedamage products in fish compared to previous reports (Tsangariset al., 2011a,b). Interestingly, the corresponding substratesarachidonic acid, eicosapentaenoic acid, docosahexaenoic acidand cholesterol in fish decreased in the presence of hydrogenperoxide. Altogether reduction of fatty acid and cholesterol, andaugmented oxidized lipid products questions quality of fish oil infish exposed to excessive oxidants (natural or man-made) forhuman consumption.

Increased oxidative stress in fish may also affect the generalhealth and growth like human and rodents (Livingstone, 2001). Itis suggested, exogenous stress to environmental change on fishmay cause energy partitioning of nutrient storage and mobiliza-tion, and likelihood to suppress reproduction and skeletal and soft

Fig. 5. Oxidized lipid products produced via lipoxygenase enzyme released by male and female medaka with or without hydrogen peroxide exposure. Significant difference⁄p < 0.05 and ⁄⁄p < 0.01 by analysis of variance between time points. All bar column values are mean ± SD, n = 3. Significant difference between 200 and 1000 lM at respectivetime are annotated by +p < 0.05 and ++p < 0.01.

M.L.S. Chung et al. / Food and Chemical Toxicology 53 (2013) 205–213 211

Author's personal copy

tissue growth (Leatherland and Woo, 2010). Information on therelationship between oxidative stress, health, disease and fitnessin aquatic organisms from either clean or pollute environmentsis limited. It is known marine fish exposed to hypoxic conditiondue to natural environmental changes such as eutrophicationundergoes physiological adjustments with an attempt to maintainhomeostasis (Cadenas, 1989, Chung-Davidson et al., 2008); how-ever, if hypoxic state persists or beyond their tolerance limit,impairment of growth and reproduction, disease susceptibility in-creases and in extreme case, death may occur. It is evident that li-pid metabolism in fish is impaired by oxidative stress. As suggestedby Leatherland and Woo (2010), fish under exposure of toxicantsor infectious agents will alter homeostasis and will either have a

compensating response (within tolerance change) and re-establishhomeostasis or have compensating response (beyond tolerancerange) with subsequent cellular dysfunction (such as DNA) andeven death. Perhaps this is so from the observation in this study.Longer exposure of hydrogen peroxide did not necessarily lead todeath (data not shown) or increase oxidized lipid products in fishand instead decreased: perhaps a natural mechanism to retainhomeostasis.

Isoprostanes are associated to many human diseases such asischemic stroke, cancer and Parkinson’s disease (Lee et al., 2009;Lim et al., 2010). In human, high concentrations of isoprostanesare associated to many vascular related diseases and neurodegen-eration (Lee et al., 2009) but it is not known if this is so in fish. BothF2-IsoP and F3-IsoPs are known to be vasoconstrictors and F4-NP tobe concentrated in neuronal membranes (Roberts et al., 1998).HETEs, COPs and resolvin D1 are also classified as lipid mediatorswhere these compounds take part in the cascade of oxidized lipidproducts in vivo. It is shown HETEs compounds are elicited in hu-man cancer cells and take part in vascular homeostasis, lung andrenal vasculature, inflammation, and renal functions (Mallatet al., 1999; Carroll and McGiff, 2000 Unterwurzacher et al.,2008; Moreno, 2009). Several of the COPs have been associatedin diseases such as coronary artery disease, dengue, and Parkin-son’s disease (Lee et al., 2009; Garenc et al., 2010) and resolvinD1 is suggested to be potent anti-inflammatory lipid mediatorsand take part in suppressing oxidative stress related to neurotox-icity (Bazan et al., 2011).

Investigations are scarce on oxidative damage biomarkers inmarine fish. Isoprostanes have been detected in gut of Atlantic sal-mon (Salmo sala L.) under acute stress but the functional role wasnot explained (Olsen et al., 2012). F2-IsoPM1 and F2-IsoPM2,metabolites of F2-IsoPs were exquisitely high in the body muscleof medaka in this study. The rationale for this cause needs to befurther elucidated. It is shown exposure of oxidants for 10–60 min showed structural and functional changes in fish gill cellsand 12 lipoxygenase (precursor of 12-HETE) was found augmentedin presence of low dose reactive oxygen species in fish gill butinhibited at high dose (German and Hu, 1990). However, theauthors did not explain the exact mechanisms on functional orphysical role for this observation. Furthermore, similar findingwas not made in this study and instead, oxidation via lipoxygenaseor cytochrome P450 produced approximately 3 times more metab-olites in fish muscle than those produced by non-enzymatic path-way. Moreover resolvin D1 from DHA in fish was recentlyidentified in Atlantic salmon (S. salar L.) but was at trace level;regardless the functional role as lipid mediator still need to beidentified (Olsen et al., 2012).

5. Conclusion

This study is the first to elucidate a range of oxidized lipid prod-ucts in marine fish under oxidative stress. Our observations in thisstudy address few factors regarding fish physiological response toexogenous oxidative stress. Response of antioxidant enzyme activ-ity is rapid but short and instead become suppressed if exposuretime is too long. As previously suggested in human studies(Halliwell and Lee, 2010), timing of sampling after a treatment isimportant. Some fish may have faster response to reactive oxygenspecies to release oxidized lipid products (isoprostanes, neuropros-tanes, 5(S)-HETE, 20-HETE, 7-ketocholesterol, 27-hydroxycholes-terol and resolvin D1), some may take time (12(S)-HETE) andsome are even suppressed (7b-hydroxycholesterol). Altogether,our findings showed oxidized lipid products in fish are generatedrapidly when under oxidant insult. It also suggests that dependingon the concentration and length of exposure time different

Fig. 6. Levels of antioxidant enzyme activities of male and female medaka with orwithout hydrogen peroxide exposure. Significant difference ⁄p < 0.05 and ⁄⁄p < 0.01by analysis of variance between time points. All bar column values are mean ± SD,n = 3. Significant difference between 200 and 1000 lM at respective time areannotated by +p < 0.05 and ++p < 0.01.

212 M.L.S. Chung et al. / Food and Chemical Toxicology 53 (2013) 205–213

Author's personal copy

biomarkers should be considered to assess oxidative stress inmarine fish. As HETEs, resolvin D1 and certain cholesterol oxida-tion products depend on the enzyme activity (lipoxygenase andcytochrome P450) there may be rate-limiting factor in the kineticsof the enzymes involved. Nevertheless, it is important to note,many of these biomarkers are related to lipid mediators ofdiseases thus information obtained is important on health statusof the fish. It is probably an advantage to rely on biomarkersnamely isoprostanes, neuroprostanes, 7-ketocholesterol and 7b-hydroxycholesterol to assess oxidative stress damage, as thereare metabolites independent from enzyme activities. Using theseproposed biomarkers further studies are required to find the effectof oxidants at low dose (at level of natural marine environment)and longer exposure time to identify the long-term effect.

Conflict of Interest

The authors declare that there are no conflicts of interest.

Acknowledgement

The authors would like to thank The University of Hong Kong,University Research Committee (Seed Funding Programme for Ba-sic Research) Project No. 201110159009 for the support of thisstudy. Mr. Robert Chui for his assistance in antioxidant enzymeassay.

References

Basu, S., 1999. Oxidative injury induced cyclooxygenase activation in experimentalhepatotoxicity. Biochem. Biophys. Res. Commun. 254, 764–767.

Bazan, N.G., Molina, M.F., Gordon, W.C., 2011. Docosahexaenoic acidsignololipidomics in nutrition: significance in aging, neuroinflammation,macular degeneration, Alzheimer’s, and other neurodegenerative diseases.Annu. Rev. Nutr. 31, 321–351.

Cadenas, E., 1989. Biochemistry of oxygen toxicity. Annu. Rev. Biochem. 58, 79–110.Carroll, M.A., McGiff, J.C., 2000. A new class of lipid mediators: cytochrome P450

arachidonate metabolites. Thorax 55, S13–S16.Chung-Davidson, Y.W., Rees, C.B., Bryan, M.B., Li, W., 2008. Neurogenic and

neuroendocrine effects of goldfish pheromones. J. Neurosci. 28, 14492–14499.Fahy, E., Subramaniam, S., Murphy, R.C., Nishijima, M., Raetz, C.H., Shimizu, T., et al.,

2009. Update of the LIPID MAPS comprehensive classification system for lipids.J. Lipid Res. 50, S9–S14.

Garenc, C., Julien, P., Lev, E., 2010. Oxysterols in biological systems: thegastrointestinal tract, liver, vascular wall and central nervous system. FreeRadical Res. 44, 47–73.

German, J.B., Hu, M.L., 1990. Oxidant stress inhibits the endogenous production oflipoxygenase metabolites in rat lungs and fish gills. Free Radical Biol. Med. 8,441–448.

Halliwell, B., Gutteridge, J.M.C., 2007. Free radicals in biology and medicine, fourthed. Oxford University Press, Oxford, United Kingdom.

Halliwell, B., Lee, C.Y.J., 2010. Using isoprostanes as biomarkers of oxidative stress:some rarely considered issues. Antioxid. Redox Signal. 13, 145–156.

Kadiiska, M.B., Gladen, C.C., Baird, D.D., Germolec, D., Graham, L.B., Parker, C.E., et al.,2005. Biomarkers of oxidative stress study II. Are oxidation products of lipids,proteins, and DNA markers of CCl4 poisoning? Free Radical Biol. Med. 38, 698–710.

Knight, J.A., Pieper, R.K., McClellan, L., 1988. Specificity of the thiobarbituric acidreaction: its use in studies of lipid peroxidation. Clin. Chem. 34, 2433–2438.

Leatherland, J.F., Woo, P.T.K., 2010. Fish Diseases and Disorders, vol. 2, CABInternational Canada.

Lee, C.Y.J., Huang, S.H., Jenner, A.M., Halliwell, B., 2008. Measurement of F2-isoprostanes, hydroxyeicosatetraenoic products, and oxysterol from a singleplasma sample. Free Radical Biol. Med. 44, 1314–1322.

Lee, C.Y.J., Jenner, A.M., Halliwell, B., 2004. Rapid preparation of human urine andplasma samples for analysis of F2-isoprostanes by gas chromatography-massspectrometry. Biochem. Biophys. Res. Commun. 320, 696–702.

Lee, C.Y.J., Seet, R.C.S., Huang, S.H., Long, L.H., Halliwell, B., 2009. Different patternsof oxidized lipid products in plasma and urine of dengue fever, stroke andParkinson’s disease patients: cautions in the use of biomarkers of oxidativestress. Antioxid. Redox Signal. 11, 407–420.

Lesser, M.P., 2006. Oxidative stress in marine environments: biochemistry andphysiology ecology. Annu. Rev. Physiol. 68, 253–278.

Liang, Y., Ping, W., Duke, R.W., Reaven, P.D., Harman, S.M., Cutler, R.G., et al., 2003.Quantification of 8-iso-prostaglandin-F2a and 2,3-dinor-8-iso-prostaglandin-F2a in human urine using liquid chromatography-tandem mass spectrometry.Free Radical Biol. Med. 34, 409–418.

Lim, K.H.C., Lee, C.Y.J., Earnest, A., Seet, R.C.S., Halliwell, B., 2010. Does radiotherapyincrease oxidative stress? A study with nasopharyngeal cancer patientsrevealing anomalies in isoprostanes measurements. Free Radical Res. 44,1064–1071.

Livingstone, D.R., 2001. Contaminant-stimulated reactive oxygen speciesproduction and oxidative damage in aquatic organisms. Mar. Pollut. Bull. 42,656–666.

Mallat, Z., Nakamura, T., Ohan, J., Lesèche, G., Tedgui, A., Maclouf, J., 1999. Therelationship of hydroxyeicosatetraenoic acids and F2-isoprostanes to plaqueinstability in human carotid atherosclerosis. J. Clin. Invest. 103, 421–427.

Moreno, J.J., 2009. New aspects of the role of hydroxyeicosatetraenoic acids in cellgrowth and cancer development. Biochem. Pharmacol. 77, 1–10.

Olsen, R.E., Svardal, A., Eide, T., Wargelius, A., 2012. Stress and expression ofcyclooxygenase (cox1, cox2a, cox2b) and intestinal eixosanoid in Atlanticsalmon, Salmo salar L. Fish Physiol. Biochem. 38, 951–962.

Quinn, J.F., Raman, R., Thomas, R.G., Yurko-Mauro, K., Nelson, E.B., Van Dyck, C.,et al., 2010. Docosahexaenoic acid supplementation and cognitive decline inAlzheimer disease. A randomized trial. JAMA 304, 1903–1911.

Roberts, L.J., Montine, T.K., Markesbery, W.R., Tapper, A.R., Hardy, P., Chemtob, S.,et al., 1998. Formation of isoprostane-like compounds (neuroprostanes) in vivofrom docosahexaenoic acid. J. Biol. Chem. 273, 13605–13612.

Roberts, L.J., Moore, K.P., Zachert, W.E., Oates, J.A., Morrow, J.D., 1996. Identificationof the major urinary metabolite of the F2-isoprostane, 8-iso-prostaglandin F2a,in humans. J. Biol. Chem. 271, 20617–20620.

Roberts, L.J., Morrow, J.D., 2000. Measurement of F2-isoprostanes as an index ofoxidative stress in vivo. Free Radical Biol. Med. 28, 505–513.

Shui, G., Cheong, W.F., Happar, I.A., Hoi, A., Xue, Y., Fernandis, A.Z., et al., 2011.Derivatization-independent cholesterol analysis in crude lipid extracts by lipidchromatography/mass spectrometry: applications to a rabbit model foratherosclerosis. ⁄⁄⁄J. Chromatography A 1218, 4357–4365.

Skulas-Ray, A.C., Kris-Etherton, P.M., Harris, W.S., Vanden Heuvel, J.P., Wagner, P.R.,West, S.G., 2011. Dose-response effects of omega-3 fatty acids on triglycerides,inflammation, and endothelial function in healthy persons with moderatehypertriglyceridemia. Am. J. Clin. Nutr. 93, 243–252.

Streppel, M.T., Ocke, M.C., Boshuizen, H.C., Kok, F.J., Kromhout, D., 2008. Long-termfish consumption and n-3 fatty acid intake in relation to (sudden) coronaryheart disease death: the Zutphen study. Eur. Heart J. 29, 2024–2030.

Tsangaris, C., Vergolyas, M., Fountoulaki, E., Nizheradze, K., 2011a. Oxidative stressand genotoxicity biomarkers responses in Grey Mullet (Mugil cephalus) from apolluted environment in Saronikos Gulf, Greece. Arch. Environ. Contam. Toxicol.61, 482–490.

Tsangaris, C., Vergolyas, M., Fountoulaki, E., Goncharuk, V.V., 2011b. Genotoxicityand oxidative stress biomarkers in Carassius gibelio as endpoints for toxicitytesting of Ukrainian polluted river waters. Ecotoxicol. Environ. Saf. 74, 2240–2244.

Unterwurzacher, I., Koal, T., Bonn, G.K., Weinberger, K.M., Ramsay, S.L., 2008. Rapidsample preparation and simultaneous quantitation of prostaglandins andlipxoygenase derived fatty acid metabolites by liquid chromatography-massspectrometry from small sample volumes. Clin. Chem. Lab. Med. 46, 1589–1597.

Valavanidis, A., Vlahogianni, T., Dassenakis, M., Scoullos, M., 2006. Molecularbiomarkers of oxidative stress in aquatic organisms in relation to toxicenvironmental pollutants. Ecotoxicol. Environ. Saf. 64, 178–189.

Van der Meij, B.S., Langius, J.A.E., Smit, E.F., Spreeuwenberg, M.D., von Blomberg,B.M.E., Heijboer, A.C., et al., 2010. Oral nutritional supplements containing (n-3)polyunsaturated fatty acids affect the nutritional status of patients with stage IIInon-small cell lung cancer during multimodality treatment. J. Nutr. 140, 1774–1780.

M.L.S. Chung et al. / Food and Chemical Toxicology 53 (2013) 205–213 213